Non-small cell lung cancer (NSCLC) is the second most common cancer and the leading cause of cancer- related deaths in this country. In general, those without spread of cancer to mediastinal lymph nodes undergo curative lung resection, and those with mediastinal lymph node disease undergo palliative chemoradiation therapy. Non-invasive and/or invasive diagnostic modalities are used to stage the mediastinum. Because of the limited diagnostic accuracy of any single modality, optimal mediastinal staging likely requires the use of more than one diagnostic modality. However, an evidence-based approach to staging has yet to be established and practice guidelines do not provide consistent recommendations. How the community at large utilizes combinations of staging modalities is unknown. Variation in the approach to staging may occur over time, geographic location, and sociodemographic factors as a result of new staging technology, scientific uncertainty, and disparities in care, respectively. Variation in the use of combinations of staging modalities has never been described. Staging likely affects outcomes through the appropriate and inappropriate allocation of treatment. Patients with understaged mediastinal lymph nodes are unnecessarily exposed to the risks and consequences of a lung resection, whereas those with overstaged mediastinal lymph nodes are denied a potentially curative therapy. An association between staging and survival has never been reported. The overall goal of this project is to characterize mediastinal staging in a cohort of patients diagnosed with NSCLC between 1998 and 2002 using a large cancer registry linked to Medicare claims. We aim to 1) describe temporal trends, regional variation, and patient factors associated with the use of different combinations of mediastinal staging modalities; and 2) explore the association between different combinations of mediastinal staging modalities and survival. Poisson regression will be used to evaluate the frequency of use of positron emission tomography (PET) over time. Logistic regression will be used to evaluate whether geographic area and/or patient characteristics predict the use of computed tomography, PET, and mediastinoscopy in combination. Finally, Cox multivariate regression will be used to perform an exploratory analysis of an association between the use of extensive mediastinal staging and survival. Findings from this work will help justify prospective clinical trials better suited to provide unbiased estimates of risk-adjusted survival for differing staging strategies. The body of evidence emerging from this line of work also has implications for quality improvement initiatives and incorporating standardized staging into clinical trials of therapy. [unreadable] [unreadable] [unreadable] [unreadable]